Papers on Parliament No. 59
April 2013
Christopher Kam and Faruk Pinar 'Paying for Parliament: Do We Get What We Pay For? Lessons from Canada*
Prev | Contents | Next
I. Introduction
The quality of a representative democracy depends on electing highly qualified and talented people to office, and then effectively motivating those people once they are in office and beyond the immediate reach of voters. The remuneration paid to office-holders presumably plays an important role in this regard. Attractive levels of remuneration make it more likely that qualified and talented individuals will seek election and that incumbents will remain in office long enough to develop valuable expertise and skills. Conversely, levels of remuneration that are too low may invite ill-equipped or ill-suited individuals to seek office and force otherwise talented office-holders to leave politics in search of more lucrative employment.
Academic observers and parliamentarians have long complained that the Canadian Parliament suffers from the latter condition. A central weakness of the Canadian House of Commons, Charles Franks laments, is its amateurism[1]: MPs are not as qualified as might be desired, and too many arrive in the House too late in life to acquire the experience necessary to act as effective legislators and representatives. A series of policies—notably the introduction of salaries and pensions for M
Ps in 1953 and substantial increases in salaries in 1963 and 2001—were adopted precisely to attract and retain a ‘better grade’ of MPs. The hope was that the increased compensation would induce talented (and well-paid) individuals to enter politics and establish long-standing parliamentary careers. We focus in this paper on the first of these dimensions, that is, the attraction and election of qualified individuals to political office (leaving the matter of retention to another paper). Specifically, we pose three questions. First, did these changes attract a younger set of parliamentary candidates or result in the election of younger MPs to the House of Commons? Second, did these same changes alter the professional backgrounds from which parliamentary candidates were drawn? Third, did the introduction and subsequent increases in MPs’ salaries and pensions encourage more highly educated candidates to stand for election to the House of Commons? Our analysis of the MPs and candidates at federal elections from 1867 to 2011 reveals little evidence that these reforms achieved their stated objectives. The candidates who contested parliamentary elections following these reforms were not younger than expected given contemporaneous changes in the general population. Similarly, whilst the introduction of salaries and pensions in 1953 and the salary increase of 1963 coincided with an increase in the number of candidates drawn from law (a highly paid field) relative to the number of candidates drawn from teaching (a modestly paid field), the increases were small in magnitude, short-lived, and statistically indistinguishable from random variation in the data. Lastly, we find no evidence that the 1953 and 2001 changes to MPs’ remuneration succeeded in increasing the percentage of university degree-holders entering the House of Commons. Only the 1963 salary increase does appear to have succeeded in this respect, however.
Our work contributes to a growing body of literature on the desirability and effectiveness of using monetary compensation to motivate and control elected representatives. Many of the theoretical contributions to this literature make clear that increasing the remuneration of elected politicians does not necessarily improve the quality of elected politicians, and it may actually produce unintended side effects. This can happen for a variety of reasons, but the literature highlights three main possibilities. First, if aspects of a politician’s job are undesirable, highly talented individuals may leave politics to less talented individuals even as the salaries attached to political office increase.[2] Second, whilst higher salaries may induce highly qualified individuals to enter and remain in politics, they also provide strong incentives for poorly qualified individuals to enter politics.[3] At the limit, poorly qualified candidates may crowd out well-qualified candidates. Third, the marginal effect of increased pay on the behaviour of politicians may diminish and sharpen the trade-offs confronting voters as it does so.[4]
These theoretical results have motivated empirical work on the remuneration of politicians. Some of this work supports the conventional wisdom that higher pay improves the quality of candidates and the performance of incumbent politicians. Claudio Feraz and Frederico Finan, for example, study the impact of federally imposed, population-based limits on the salaries of Brazilian municipal politicians.[5] They find significant jumps in the quality, productivity, and tenure of Brazilian municipal politicians at the boundaries of each salary band. In an effort that is similar to our own, Michael Atkinson and Dustin Rogers show that a 2001 increase in Canadian MPs’ salaries resulted in the election of more educated MPs at the subsequent election in 2004.[6]
Other empirical studies, however, demonstrate that some of the more pessimistic predictions of the theoretical models are also encountered in the real world. Kaisa Ilona Kotakorpi and Panu Poutvaara, for example, show that a significant increase in the remuneration of Finnish MPs led to an increase in the educational status of female candidates but had no discernible effect on the educational status of male candidates.[7] The result demonstrates higher pay may not always or uniformly improve the quality of candidates who offer themselves for public office. Raymond Fisman et al. demonstrate a more worrisome possibility.[8] The authors consider the impact of a 1994 harmonisation of the salaries of members of the European Parliament (MEPs). Prior to harmonisation, the salaries of MEPs varied substantially depending on the MEP’s nationality. Fisman et al. show that MEPs for whom harmonisation implied a salary increase were more likely to run for re-election than MEPs for whom it implied a salary decrease.[9] However, they also find that the overall quality of MEPs as measured by their educational pedigree declined following harmonisation. This last result is consistent with the theoretical prediction that increased political remuneration may lead low-quality candidates to crowd out high-quality candidates.
We adopt an approach that is common in many of these empirical studies, and take advantage of a series of policy interventions (specifically the 1953 introduction of salaries and pensions and salary increases in 1963 and 2001) to identify the empirical implications of increasing the remuneration offered to Canadian MPs. We also share with a number of these studies a focus on the educational attainment of candidates. Educational attainment is an observable correlate of political skill (e.g., of rhetorical and analytical ability), but it is also a proxy for an individual’s earning power in the outside economy and hence of their opportunity cost for participating in politics. An increase in MPs’ salaries and pensions should therefore attract more educated candidates to contest elections. We focus on candidates’ professional backgrounds (and the teaching and legal professions in particular) for similar reasons. Not only are many more candidates drawn from these two professions than any others, but we have good information from the census on the percentage of teachers and lawyers in the general labour force and on the average earnings of individuals in these two professions. This allows us to assess whether MPs’ salaries and pensions were sufficient to encourage lawyers and teachers to become parliamentary candidates in greater numbers than we might expect given the prevalence of these professions in the labour force. Finally, our interest in the age distribution of parliamentary candidates stems from longstanding complaints about the amateurism of the Canadian MPs. If higher levels of remuneration do not induce a younger set of parliamentary candidates to contest elections, and do not result in the election of younger MPs, they are very unlikely to generate longer parliamentary careers. Our examination of the age of parliamentary candidates and first-time MPs sets our study apart from Atkinson and Rogers’ study.[10] However, we would also emphasise that our study is also more extensive that Atkinson and Rogers’ study (which focuses on the period between 2000 and 2004) in that it takes in a much longer span of Canada’s post-confederation history (we trace changes in MPs’ remuneration from 1867).
The remainder of the paper is structured as follows. The section immediately following this introduction sets out the context in which the reforms to MPs’ salaries and pensions were introduced. We begin by describing the House of Commons’ amateurism problem in greater detail. We then relate the grounds on which policy-makers justified the introduction (and increases) of parliamentary salaries and pensions. This section makes clear that a central motivation for the introduction and increase in MPs’ salaries and pensions was to induce younger and more qualified individuals to enter the House of Commons. Implicit in the argument that higher levels of remuneration were needed to attract more qualified individuals to the House of Commons is the claim that existing levels of remuneration were insufficient to achieve this. Thus we end the section by describing the history of parliamentary remuneration in Canada and making an effort to develop some standardised basis on which to assess trends in the economic value of MPs’ earnings over time. The third section of the paper considers whether the introduction of MPs’ salaries and pensions in 1953 and the substantial increase in salaries in 1964 and 2001 resulted in a more professionalised, educated, and younger pool of candidates and first-time MPs. The fourth section of the paper discusses our results.
II. Context
1. The amateurism of the Canadian House of Commons
Michael Atkinson and David Docherty observe that Franks’[11] description of Canadian MPs as political amateurs might be interpreted in a number of ways.[12] Amateurism could be taken to imply a lack of training or dilettantism, for example. Certainly, this aspect of Canadian MPs’ amateurism is recognised in the literature. ‘The most notable feature of the Canadian data on previous elected experience,’ V.S. Harder wrote, ‘is the overwhelming lack of it’.[13] However, Franks’ description of Canadian MPs as amateurs was also intended to stress that they were not career politicians.[14] The latter label was due to Anthony King, who used it to describe the character of typical British MPs of the postwar period.[15] British MPs were career politicians not so much by virtue of the fact that they earned their incomes from politics as by their deep psychological commitment to life in politics. These were individuals who, as Weber phrased it, did not just ‘live off politics, but for politics.’[16]
One of the empirical hallmarks of a career politician, King argued, was the early age at which they entered the House of Commons.[17] Early entry to the House was important for two reasons. First, it was almost a necessary initial condition of a long parliamentary career. Second, an early entry into politics implied that the individual had surrendered outside career options (in law, medicine, business, etc.) and thereby fully (if implicitly) committed themselves to a political career. Franks’ description of Canadian MPs’ as political amateurs thus implies a corresponding lack of commitment to a political career. Many Canadian MPs perceived their job as a form of community service that was ancillary to their established professional careers. Canadian MPs therefore entered the House later in life, their time in politics representing either a sabbatical from or a capstone to their professional careers. Indeed, prior to 1900 (in the days when the MP’s job was a part-time one) the average age of Canadian MPs on their first entry to the House was 42 years; by 1950 the average age of entry had actually increased to 44 years, despite the fact that the MP’s job was by that time a full-time one. Identifying whether the introduction and increase in salaries and pensions lowered the average age at which first-time MPs entered the House is, of course, one of our central concerns.
2. Arguments for higher salaries and pensions
Amateurism is an academic label, and MPs themselves did not employ the term when they debated the introduction and subsequent adjustment of salaries and pensions. It is clear from their comments, however, that MPs had arrived at a similar diagnosis of the situation. When the prime minister of the day (Rt. Hon. Louis St. Laurent) introduced the legislation to provide MPs with salaries and pensions, he justified the measure as one that would ‘strengthen our parliamentary institutions’. Salaries and pensions, St. Laurent argued, ‘would encourage the right kind of men to run for public office, and afterwards, when they had acquired experience, to remain in public life as long as their constituents wished them to do so’.[18] Implicit in St. Laurent’s comments was the contention that parliament had failed to attract and retain the ‘right kind of men’.[19] Agar Adamson, who followed St. Laurent in the debate, was more direct. Professionals could not be expected to give up their careers and enter the House under the current arrangement (i.e., the sessional indemnity), Adamson argued, ‘A pension scheme was necessary if the House was to induce professionals to become MPs’.[20]
Similar arguments were presented by the government house leader (Rt. Hon. Mitchell Sharp) when the issue of parliamentary salaries and pensions was revisited in 1974. More important than the financial pressure that low salaries placed on the House’s current members was that, ‘people who should be considering running for Parliament are deterred by the insecurity coupled with the low prevailing rates of remuneration’.[21] Gerald Baldwin spoke for the official opposition, and he largely concurred with Sharp’s assessment. It was imperative, Baldwin stressed, that the House be provided with ‘a continuing inflow of people—hopefully young people [our emphasis]’.[22] An inflow of youth was vital because it took many years for an individual to develop into an effective member.
Concerns about the financial pressures on sitting MPs were also expressed in these debates, but as this brief review makes clear, leading parliamentary figures were at least as concerned with the effect of salaries and pensions (or the lack thereof) on the type of people who sought entrance to the House of Commons. The argument was repeated in periodic reports on the remuneration of MPs. Higher pay was necessary to attract young professionals ‘of proven ability’[23], and the plain fact, as one report directly stated, was that ‘good pay is required to attract good people’.[24] Of course, not all MPs who participated in these debates agreed with the adoption (and later increases) of salaries and pensions. Most objections, however, were motivated by a concern for economy or to avoid the appearance of self-dealing. In contrast, few opponents of these measures disputed the central claim that the levels of parliamentary remuneration then on offer made it difficult to attract and retain talented individuals to the House of Commons.[25]
3. The remuneration of MPs, 1867–2011
Implicit in the argument that higher salaries and pensions were needed to attract young and highly capable people to the House of Commons is the claim that existing levels of parliamentary remuneration were insufficient to do this. Judging the factual basis of this claim requires examining what MPs earned at various points in time and developing some basis for comparing the value of those earnings over time. We begin this task by relating the history of remuneration of the members of the House of Commons. This history can be divided into three broad periods:
- An initial period (1867–1952) where the main vehicle of remuneration was a sessional indemnity;
- A second period (1953–2000) which saw the introduction of pensions and annual salaries supplemented by a series of allowances;
- Finally, a third period (2001–2012) in which the elimination of allowances was offset by a substantial increase in MPs’ salaries and (less noticed) an increase in the set of parliamentary positions that drew additional stipends. Elements of the pension plan were also scaled back during this third period.
Table 1 below offers further details on the amount and nature of the remuneration provided to Canadian MPs in each of these periods. The information for the years benchmarking the beginning and end of each period provide a sense of the pace and nature of change in MPs’ pay over time. Changes in the main instruments of remuneration can be understood as institutional responses to broader political forces. The switch in 1953 from a system of sessional indemnities and per diem allowances to one of fixed annual salaries and pensions, for example, reflected the professionalisation of Canadian politics and a shared recognition among MPs that national politics no longer operated on a part-time or casual basis. Similarly, the consolidation of MPs’ various allowances into a single parliamentary salary in 2001 was a concession to the public’s demand for greater transparency in how MPs were remunerated.
It is hard to compare the economic value of MPs’ remuneration at different points of time on the basis of table 1. First, the full value of MPs’ remuneration prior to 1953 hinged on the number and duration of parliamentary sessions in any given year. Second, the nominal amounts of MPs’ non-taxable allowances understates the full economic value of these allowances. Third, the amounts shown in table 1 are expressed in nominal terms and so do not reflect changes in purchasing power or relative earning power over time. Figures 1–3 attempt to put the amounts shown in table 1 on a more comparable footing. To do this, we have:
- translated the sessional indemnities and per diem allowances of the first period into annual amounts conditional on the number and length of the parliamentary session in each year;
- translated MPs’ non-taxable expense allowances into a taxable equivalent and added this amount to their base salaries;[26] and
- translated the resulting amounts into real terms on the basis of i) the consumer price index (so as to account for inflation and control for purchasing power); and ii) average salaries in law and teaching, two professions from which many MPs are drawn (so as to account for relative earning power and give a sense of the opportunity cost of serving in the House of Commons).
Table 1: The remuneration of Canadian MPs, 1867–2012
|
|
Main vehicle of remuneration
|
Pension
|
Additional allowances
|
|
Period 1
(1867–1952)
|
Sessional indemnity
|
|
|
|
1867
|
· $600 for sessions >30 days
· $6 per diem for sessions < 30 days
|
―
|
· Mileage compensated at $0.10 per mile
|
|
1952
|
· $4000 for sessions > 65 days;
· $25 per diem for sessions < 65 days
|
―
|
· $2000 tax-free expense allowance
|
|
Period 2
(1953–2000)
|
Salary
|
|
|
|
1953
|
· $8000 per annum
|
· Vested after: 6 years
· Age of eligibility: on defeat or resignation
· Indexed to CPI at age 60
· Accrual rate: 5%
· Limit: 75% of best 6 years’ salary
· MP’s contribution: 11% of salary
|
· $2000 tax-free expense allowance
|
|
2000
|
· $68 200 per annum
|
· Vested after: 6 years
· Age of eligibility: 55
· Indexed to CPI at age 60
· Accrual rate: 4%
· Limit: 75% of best 5 years’ salary
· MP’s contribution: 9% of salary
|
· $22 500 tax-free expense allowance
· $12 000 additional (housing) allowance
|
|
Period 3
(2001–2012)
|
Salary
|
|
|
|
2001
|
· $131 400 per annum
|
· Vested after: 6 years
· Age of eligibility: 55
· Indexed to CPI at age 60
· Accrual rate: 3%
· Limit: 75% of best 5 years’ salary
· MP’s contribution: 7% of salary*
|
N/A—but many parliament positions draw additional stipends
|
|
2012
|
· $157 731 per annum
|
Source: Indemnities, Salaries and Allowances: Members of the House of Commons (www.parl.gc.ca/ ParlInfo/Lists/Salaries.aspx?Menu=HOC-Politic&Section=03d93c58-f843-49b3-9653-84275c23f3fb). *The rate of MPs’ contributions has varied over time.
Thus, we ignore the economic value of mileage and housing allowances and pensions. The former can safely be ignored because the amounts are generally small. Excluding pensions from the calculation is more controversial, however, because these amounts can be quite significant. For example, any current MP who retired at 60 years of age after just eight years’ service would be entitled to upwards of $50 000 per annum indexed to inflation—and this sum might well be collected for 20 years given modern lifespans. This substantially distorts any effort to compare the financial position of MPs before and after the introduction pensions. Equally, any effort to translate pensions into an income equivalent rests on a host of actuarial assumptions (e.g., the average length of an MPs’ tenure, the probability that the MP held a parliamentary position that conferred an additional stipend, the average age of retirement and years of survival after retirement, etc.) that inject tremendous uncertainty into any resulting estimate. More important, we feel, is to recognise that the introduction of pensions fundamentally reduced the risk inherent in a parliamentary career.
Figure 1: MPs’ annual remuneration in real terms, 1867–2011
Sources: Data on MPs’ remuneration obtained from Indemnities, Salaries and Allowances: Members of the House of Commons (www.parl.gc.ca/ParlInfo/Lists/Salaries.aspx?Menu=HOC-Politic&Section= 03d93c58-f843-49b3-9653-84275c23f3fb). Price indexes obtained from Blais Commission, Supporting Democracy (1998, vol. 2, pp. 88–91) and CANSIM Table: 3260020 Vector: V41690973.
Figure 1 shows a long-run increase in remuneration of Canadian MPs in real terms. Certainly, there were identifiable periods of stagnation in the level of remuneration that MPs’ received, for example, between 1955 and 1963 and 1975 to 2001. Equally, pay increases in 1964 and 2001 more than offset any short-term decline in the purchasing power of MPs’ salaries. Even the decline in the real value of MPs’ remuneration immediately following the introduction of salaries is a bit of an illusion. When MPs were paid by the session, years in which the House sat for two sessions brought MPs a windfall. Such years were rare in the 1920s and 1930s, and hence the $8000 salary introduced in 1953 significantly increased the level of MPs’ remuneration relative to the 1920s and 1930s. In 1949, 1950, and 1951, however, there were two sessions in the year and MPs earned $8000 per annum. The initial annual salary of $8000 thus appears as a continuation of immediately prevailing conditions, but these conditions were quite atypical. Indeed, the overall upward trend in figure 1 belies the claim (oft-repeated in the Hansard and various commission reports) that the real value of MPs’ remuneration was in long-term decline; quite the opposite, the upshot of figure 1 is that the real value of the remuneration paid to MPs was increasing over time. The central reason why our estimates show no evidence of a long-term decline in the real value of MPs’ remuneration is that we have included the non-accountable, tax-free allowances that MPs received between 1945 and 2000 as part of MPs’ annual earnings. The decision is easy to defend: the Blais Commission report itself remarked that MPs had come to view their tax-free allowances as part of their salaries.[27]
MPs’ earnings may have outpaced inflation, but they did not necessarily keep pace with the growing demands of the job—and these had changed dramatically over the course of 150 years. The job of a nineteenth-century Canadian MP was a part-time one. There would typically be just one session per year. Sessions comprised an average of 70 sitting days and a little over 100 days in total. There were no specialised committees to attend. By the 1950s the job of an MP was a full-time one and an exceedingly demanding one. Sessions nowadays take up the whole calendar year, requiring MPs to travel regularly between Ottawa and their constituencies and to maintain two permanent residences. In addition to participating in the chamber, MPs sit on several committees and grapple with ever more complex policy questions. MPs always had to attend to their constituents, of course, but electorates are now so large that MPs must establish and then manage an office, staff, and budget to meet constituents’ needs.
Figure 2: MPs’ real remuneration ‘annualised’ by the duration of parliamentary sessions
Source: Parliaments—Duration of Sessions, www.parl.gc.ca/ParlInfo/compilations/parliament/ Sessions.aspx?Language=E.
To better compare the value of MPs’ pay across time, figure 2 ‘annualises’ remuneration by multiplying MPs’ earnings by a weighted inverse of the session’s duration. A session of 122 days, for example, would see that year’s pay multiplied by three (i.e., 365.25/122).[28] This methodology gives one a sense of what MPs in the era of sessional indemnities would have earned annually had their jobs been full-time ones as is the case nowadays. The Blais Commission used a very similar methodology. Estimates of the relative value of MPs’ pay generated by this methodology are certainly more supportive of the commission’s contention that the value of MPs’ pay had stagnated. There are points, notably in the late 1950s and early 1960s, where the introduction of fixed salaries left MPs worse off than they were in the late 1890s given the time demands of the job. Bear in mind that this comparison rests on the counter-factual notion of MPs in the 1870s having been paid year round; in fact, they were not. Moreover, as we noted above, the effect is amplified by the fact that MPs were unusually well-compensated for their time in the few years just prior to the introduction of salaries as a result of there being two sessions per year. Even so, it is clear from figure 2 that the 1964 pay hike was required to bring MPs’ remuneration back into line with increased time demands of the job. The 2001 pay increase had a similar albeit less dramatic effect.
Figure 3: MPs’ annual remuneration compared to salaries of lawyers and teachers and to the 99th percentile income cut-off
Sources: World Top Incomes Database (http://topincomes.g-mond.parisschoolofeconomics. eu/#Home:); Dominion Bureau of Statistics: Seventh Census of Canada, 1931 (Ottawa: J.O. Patenaude, King’s Printer, 1936); Eighth Census of Canada, 1941 (Ottawa: E. Cloutier, King’s Printer, 1943); Ninth Census of Canada, 1951 (Ottawa, E. Cloutier: Queen’s Printer, 1953); Census of Canada, 1961 (Ottawa: King’s Printer, 1967); Census of Canada, 1971 (Ottawa: Statistics Canada, 1977); Census of Canada, 1981 (Ottawa: Statistics Canada, 1983); Census of Canada, 1991 (Ottawa: Statistics Canada, 1994); 2001 Census of Population (Statistics Canada catalogue no. 97F0019XCB2001003); 2006 Census of Population (Statistics Canada catalogue no. 97-563-XCB2006062); Employment Income Statistics in Constant (2005) Dollars, Work Activity in the Reference Year, Occupation—National Occupational Classification for Statistics 2006 (720A) and Sex for the Population 15 Years and Over With Employment Income of Canada, Provinces and Territories, 2000 and 2005—20% Sample Data (2006 Census Analysis Serial no. 97-563-XCB2006062) (Ottawa: Statistics Canada, 2008).
Figure 3 offers a third perspective on the value of MPs’ remuneration over time. The figure expresses the annual salaries of lawyers and teachers and the 99th percentile income cut-off as percentages of MPs’ annual remuneration. In 1921, for example, lawyers’ average annual salaries comprised approximately 60 per cent of an MP’s annual remuneration, whilst teachers’ average annual salaries comprised just below 40 per cent. In other words, MPs were better paid in 1921 than the average lawyer and much better paid than the average teacher. Indeed, the level of MPs’ remuneration in 1921 left them just outside of the highest one per cent of income earners in the country at that time. One can discern in figure 3 a ratchet-like pattern to MPs’ relative financial positions in which they decline steadily for some period until sharply increased by a pay hike. Note, for example, that MPs did very well in the 1930s and 1940s relative to teachers, lawyers, and the very rich (i.e., the top one per cent). This was because MPs’ remuneration in those decades was conditioned on the number and length of parliamentary sessions, and hence in contrast to the incomes of other groups was largely unaffected by the long economic downturn of the Great Depression and Second World War. One can also see that the introduction of salaries in 1953 left MPs in an enviable financial position, well within the top one per cent of income earners in 1953. This was then followed by a steady decline in the relative value of MPs’ incomes as the incomes of teachers, lawyers and the very rich increased rapidly in the postwar period. The 1964 pay raise brought this decade-long decline to an end and restored MPs’ relative financial advantage. A much longer period of decline then set in. The decline accelerated in the 1990s (a product of a freeze in MPs’ pay and an economic boom that propelled incomes of the well-off) until it too was halted by the pay raise of 2001.
Despite this ratchet-like rise and fall in MPs’ relative earnings, the central thrust of figure 3 is that MPs have always enjoyed a premium in earnings relative to lawyers and teachers, at least since 1921 (which is as far back as we have been able to recover these data from the census). This is a striking result because it stands in sharp contrast to the conclusions that the Blais report drew from similar data. The Blais Commission’s data showed that lawyers were out-earning MPs by 1985 and significantly out-earning them by 1990. The same report showed that teachers’ annual salaries were on average 87 per cent of MPs’ salaries in 1996[29], much higher than any corresponding figure in our own data. These results contrast sharply with ours. A number of factors contribute to these discrepancies, but the most important is our decision to count as part of MPs’ annual remuneration the taxable equivalent of MPs’ non-accountable expense accounts. When these allowances were phased out in 2000, they amounted to $22 500 per annum. We estimate the taxable equivalent of this $22 500 to be $31 500, or about 46 per cent of an MP’s 2000 base salary of $68 200. The Blais Commission ignored these amounts and based the comparison solely on MPs’ salaries. We consider their decision an odd one because the commissioners themselves stressed that MPs had come to view the non-accountable tax-free allowance as part of their salaries.[30]
It is useful to review the discussion to this point. Salaries and pensions were introduced in 1953 because MPs largely agreed that the low pay and inherent insecurity of a parliamentary career deterred young, qualified professionals from entering the House. Pay hikes in 1964 and 2001 were predicated on similar arguments. Our examination of the level of MPs’ remuneration suggests that MPs’ earnings were never so low as to make being an MP a wholly unattractive economic proposition; over the long run MPs’ pay kept pace with inflation and with earnings in law and teaching. However, there were periods where a sharp increase in the workload relative to the pay (e.g., 1955–1963 in figure 2) or a salary freeze (e.g., 1990–2000 in figure 1) caused the value of MPs’ remuneration to stagnate temporarily. The data make clear, however, that the pay hikes of 1964 and 2001 (especially the former) brought about abrupt ends to these periods of stagnation. It is not unreasonable to hypothesise that such surges in the level of MPs’ remuneration would attract young professionals to contest parliamentary elections. It is less clear that the introduction of salaries and pensions in 1953 was as powerful an inducement to these sorts of candidates. The $8000 salary was a significant increase on what MPs earned in the 1930s (see figure 1) and it left them well within the top one per cent of Canadian income earners (see figure 3), but it may not have adequately compensated MPs for the increased time demands of the job (see figure 2). That said, the introduction of pensions ostensibly reduced the risk inherent in a parliamentary career, and this should have made it a more attractive career option for young professionals.
III. Did salaries and pensions work as intended?
1. Methodology
In this section we assess whether the introduction of and subsequent increases in MPs’ salaries and pensions induced younger, more educated, and more professional individuals to stand for office and enter the House. Our methodology is straightforward in that it involves simple comparisons of the characteristics of candidates before and after changes in level and form of MPs’ remuneration. There is always a possibility that changes in the characteristics of the candidate pool simply reflect contemporaneous changes in the population at large. To account for that possibility we focus on whether there is a discontinuity in the characteristics of the candidate pool (e.g., a dip in candidates’ average age) relative to the population at large immediately following these changes. Still, this is a lenient standard of evidence. If changes to MPs’ salaries and pensions do not generate effects that meet this standard, we have little reason to accept the broad claim on which these policies were based, i.e., that high levels of remuneration are necessary to attract highly qualified individuals to the House of Commons.
We also track changes in the age distribution, professional composition, and educational attainment of first-time MPs. This is partly because the quality of data on MPs is sometimes better than it is for candidates, but more importantly because the ultimate aim of these policies was to improve the quality of MPs in the House.[31] We caution against focusing too narrowly on the qualities of new MPs, however, as doing so can lead to misleading inferences. For example, if different parties attract candidates from different socio-economic backgrounds (and this seems likely), the characteristics of new MPs will be a function of the ebb and flow of the parties’ respective fortunes at elections rather than a response to changes in the level of MPs’ salaries.
2. Data sources
The Library of Parliament’s PARLINFO database on members of parliament and parliamentary elections and candidates serves as our main source of data.[32] One of the main strengths of the database is that it provides an occupation for almost every single candidate who has contested a parliamentary election. This enables us to track changes in the occupational composition of the set of parliamentary candidates with a fair degree of accuracy. It is sometimes the case that the same individual is listed as having more than one occupation (e.g., a lawyer who is elected as an MP is subsequently listed as a parliamentarian). In addition, not all occupational labels can be readily matched to the occupational categories employed by the census, obscuring information about an individual’s earning power in the outside economy (e.g., as when a candidate is listed as a ‘businessman’ or ‘retired’. We deal with these issues by recording the first-listed occupation for a candidate unless that occupational label is uninformative, in which case we replace it with the second-listed occupation if one is available. This mainly affects parliamentarians and retirees who we list under their prior occupations when at all possible.
The PARLINFO database is less helpful in tracking candidates’ ages and educational qualifications. For the most part, the database provides these data only for MPs. However, we are able to get a fairly good sense of the age distribution of unelected candidates by using the data provided by the three candidate surveys conducted in 1993, 2008, and 2011. We are also able to find the ages of many candidates by searching for profiles of parliamentary candidates, and these were relatively easy to find from 1984 onward. For educational qualifications, however, we are confined to the data that PARLINFO provides on MPs. The database often lists MPs’ academic and professional designations and titles (e.g., The Hon. Stephane Dion, BA, MA, PhD.), and from these designations we can accurately infer that the MP in question attained a certain level of education. We are sensitive to the fact that the absence of any such designations cannot necessarily be taken to indicate that an MP did not attain a certain level of education. When an MP’s educational credentials were not listed but the MP’s prior occupation was such that a particular level of education could be inferred safely (e.g., The Hon. Dr Grant Hill, MD), we assumed the individual had the minimum level of education consistent with that occupation. In particular, we assumed that medical doctors, university professors, lawyers, engineers, and teachers had at least an undergraduate degree. We confined these inferences to the period after 1970 by which time most professions had become credentialed and formalised. Few secondary teachers in the 1950s, for example, had university degrees; by the 1980s, most did. We were able to check some of our inferences against concrete data from the 1993 candidate survey. It was only very rarely (i.e., in fewer than one case in 100) that we imputed a higher level of educational attainment to an individual than they in fact possessed; it was more often the case that we underestimated the individual’s level of education because many of the professionals we assumed to have undergraduate degrees also held higher degrees. For this reason, we measure educational attainment by the percentage of undergraduate degree holders among MPs. We are confident that this statistic is a reliable estimator of the minimum percentage of undergraduate degree holders among MPs, but stress that it is not to be taken as a measure of the average level of education among MPs.
Finally, we can note that our data on the Canadian population (or more often the economically active portion of the Canadian population aged between 18 and 64 years) are drawn mainly from the decennial censuses conducted in the first year of each decade. Half-decennial censuses were also conducted in 1986, 1996, and 2006. We are often able to take advantage of annual statistical surveys conducted by the Dominion Bureau of Statistics or Statistics Canada to fill in some of the data in non-census years. Where no such data are available, we assume a linear interpolation of the data between census years.
We remind the reader that our interest in the age, occupational status, and educational qualifications of parliamentary candidates is not motivated by any normative contention as to the superiority of young, educated professionals as MPs. Rather we are interested in these characteristics mainly because MPs themselves saw higher salaries and pensions as means to attract people with these characteristics to the House of Commons. Our interest in the age of candidates also comes about because it is an obvious metric of the amateurism of the Canadian House of Commons. Canadian MPs are unlikely to become long-serving professional politicians unless they manage to enter the House at a young age; this requires that they stand as candidates at a young age. Similarly, it is useful to focus on these traits (as opposed to less well-defined notions of quality, competence, or talent) because they provide a concrete means to assess whether changes in the level of MPs’ remuneration had any effect. We are well aware that youth, professional standing and educational attainment are imperfect indicators of political quality, competence or talent. However, it strikes us that if changes in the level of MPs’ remuneration have no effect on the age, occupational status, and educational qualifications of parliamentary candidates (and no better indicator of quality, etc. is on hand), then one has no empirical basis whatsoever for claiming that higher pay attracts better politicians. This does not imply that evidence of such changes is sufficient to establish that higher pay attracts better politicians, but such evidence is certainly necessary if the claim is to be empirically grounded.
3. Changes in age
Figure 4 graphs the average age (in years) of non-incumbent candidates and MPs elected for the first time. Figure 4 also shows the average age of the Canadian population 18 to 64 years. One might reasonably consider this group to be the population from which candidates are drawn, and hence it provides a frame of reference with which to assess whether or not changes in the age of candidates and MPs are merely reflective of similar changes in the population at large. We have reliable data on candidates’ ages only for the 1984–2011 period, and so we focus on the impact of the 2001 pay hike. The hypothesis is that the pay hike should induce younger individuals to contest elections in an effort to enter the House because it makes a parliamentary career a relatively more attractive career option. Put differently, a significant increase in MPs’ salaries might induce those young professionals who might only enter politics once well-established in their chosen fields instead to enter politics immediately and establish political careers. Consistent with that hypothesis, figure 4 shows a marked decline in the age of non-incumbent candidates (and a correspondingly modest decline in the age of first-time MPs) at the 2004 elections, the first election held after the 2001 pay increase came into effect.[33] These changes, moreover, ran against the trend of a gradually ageing population of 18–64 year-olds.
Figure 4: Age of non-incumbent candidates and first-time MPs, 1984–2012
Sources: Candidates: 1993 Canadian Candidate Study; 2008 Comparative Candidate Study; Survey of Candidates of the 41st Federal Election (www.elections.ca/res/rec/eval/pes2011/can/can_survey_e.pdf). In addition, the authors conducted newspaper searches for candidates’ birth dates at all these elections. MPs: Members of the House of Commons (www.parl.gc.ca/parlinfo/Lists/Members.aspx); CANSIM Table 51001 (http://estat.statcan.gc.ca/cgi-win/CNSMCGI.EXE?Lang=E&RootDir=Estat/ &ESTATTemplate=Estat/CII_PICK&Version=2&Array_Pick=1&ArrayId=510001&SDDSLoc=//www.statcan.ca/cgi-bin/sdds/sdds.cgi?sdds=*).
If we confined our attention to the dramatic change in the average age of non-incumbent candidates at the 2000 and 2004 elections, we might well conclude that the 2001 pay increase worked as intended. However, this conclusion does not hold up well when considered against the full span of the data. First, the decline in the age of non-incumbent candidates in 2004 is relative to a spike in the average age of non-incumbent candidates at the 2000 election. Second, the decline in the age of non-incumbent candidates in 2004 is quickly reversed at the 2008 election. A more compelling interpretation of the data is that the trend in the average age of candidates simply mirrors that in the general population, with differences between the two groups a function of a fairly stable age difference between candidates and the population (of about seven years) and the short-run effects of election swings. The low average age of non-incumbent candidates and first-time MPs in 1984, for example, coincides with the election of a massive Progressive Conservative majority under Brian Mulroney. Similarly, the decline in the average ages of non-incumbent candidates and first-time MPs in 2011 is due mainly to the surprising electoral breakthrough of the New Democratic Party and election of many first-time NDP candidates. Consistent with this alternative interpretation is the fact that the data definitively reject the hypothesis that the average age of candidates declined after 2001; in fact, it increased by two years—exactly as much as the population average.
4. Changes in professional background
The full set of occupations from which candidates are drawn is both highly diverse and unstable over time (e.g., with telegraph operators giving way to software engineers). It is neither sensible nor informative therefore to assess changes in the occupational diversity of candidates and MPs by tracking every change in every occupation. We focus instead on changes in the ratio of lawyers to teachers among non-incumbent candidates and first-time MPs. We do so for three reasons. First, lawyers and teachers are identifiable and longstanding occupational groups. Second, law and teaching are among the largest sources of candidates and MPs, with just over 18 per cent of candidates and 30 per cent of MPs drawn from these two occupations. Third, we have a clear sense (from figure 3) of the relative opportunity costs of serving in parliament for members of these two professions. This information allows us to construct a reasonable hypothesis about how the ratio of lawyers to teachers in the candidate pool should respond to changes in the level of MPs’ remuneration.
Our theoretical argument begins with the observation that being an MP has always been a less attractive economic proposition for lawyers than teachers. It is therefore reasonable to assume that changes in the remuneration of MPs have a much greater marginal effect on the propensity of lawyers rather than teachers to seek election to parliament. Consider, for example, the hypothetical impact of a modest change in MPs’ remuneration in the 1990s when the annual incomes of lawyers and MPs were almost equal. At that point, a deterioration in the remuneration of MPs would be sufficient to dissuade the average lawyer from taking up a political career. In contrast, an improvement in the remuneration of MPs at this point might be sufficient to convince the average lawyer to enter or remain in politics. For the average teacher, however, a career as an MP implies (and has always implied) a substantial increase in income. Modest increases or decreases in the level of MPs’ remuneration do not really alter this essential fact and hence should have little impact on the average teacher’s decision to enter or withdraw from politics. If changes in MPs’ remuneration mainly affect the propensity of lawyers to enter politics and have little effect on the propensity of teachers to enter politics, it follows that the ratio of lawyers to teachers (i.e., the number of lawyers divided by the number of teachers) in the candidate pool should increase when MPs’ pay rises and decrease when MPs’ pay falls.[34]
Figure 5: The ratio of lawyers to teachers among non-incumbent candidates and first-time MPs
Sources: The national statistics such as number of lawyers, teachers and labour force were found in following resources: Dominion Bureau of Statistics, Seventh Census of Canada, 1931 (Ottawa: J.O. Patenaude, King’s Printer, 1936); Eighth Census of Canada, 1941 (Ottawa: E. Cloutier, King’s Printer, 1943); Ninth Census of Canada, 1951 (Ottawa: E. Cloutier: Queen’s Printer, 1953); Census of Canada, 1961 (Ottawa: King’s Printer, 1967); Census of Canada, 1971 (Ottawa: Statistics Canada, 1977); Census of Canada, 1981 (Ottawa: Statistics Canada, 1983); Census of Canada, 1985 (Ottawa: Statistics Canada, 1989); Census of Canada, 1991 (Ottawa: Statistics Canada, 1994); 2001 Census of Population (Statistics Canada catalogue no. 97F0019XCB2001003); 2006 Census of Population (Statistics Canada catalogue no. 97-563-XCB2006062); Employment Income Statistics in Constant (2005) Dollars, Work Activity in the Reference Year, Occupation—National Occupational Classification for Statistics 2006 (720A) and Sex for the Population 15 Years and Over With Employment Income of Canada, Provinces and Territories, 2000 and 2005—20 per cent Sample Data (2006 Census Analysis Serial no. 97-563-XCB2006062) (Ottawa: Statistics Canada, 2008).) Statistics Canada. Table 051-0001—Estimates of population, by age group and sex for July 1, Canada, provinces and territories, annual (persons unless otherwise noted), CANSIM (database), Using E-STAT. Occupational data on MPs and candidates are obtained from www.parl.gc.ca/parlinfo/Lists/Members.aspx.
Figure 5 above shows the ratio of lawyers to teachers among non-incumbent candidates and first-time MPs from 1940 to 2011. The graph also shows the ratio of lawyers to teachers in the Canadian labour force. The thick dashed line representing the latter ratio is essentially flat. This tells us that there were no fundamental changes in the balance of lawyers to teachers in the labour force throughout the period, and suggests that it is unlikely that changes in the ratio of lawyers to teachers among candidates and MPs are a result of changes in economy at large. The most striking feature of figure 5 is the sharp decline in the ratio of lawyers to teachers among first-time MPs. It is tempting to interpret this decline as evidence of lawyers withdrawing from politics in response to the decade-long stagnation of the value of MPs’ pay in the mid-1950s (per figures 1 and 2), but such an interpretation cannot be sustained. The ratio of lawyers to teachers among first-time MPs confounds the willingness of members of these two professions to enter politics with their relative success at getting elected. It is the former not the latter that we expect to be affected by changes in the value of MPs’ pay. The ratio of lawyers to teachers among non-incumbent candidates was actually quite stable from the late-1940s to the early 1960s. This suggests that lawyers remained quite willing (or at least as willing as teachers) to enter politics notwithstanding any stagnation or deterioration in the real value of MPs’ remuneration, at least until 1962. There was at that point a steep and sudden decline in the ratio of lawyers to teachers among non-incumbent candidates at the 1962 election. This decline was not due in any obvious fashion to changes in the level or value of MPs’ remuneration, however: it occurred nine years after the introduction of salaries and pensions and a year before the 1963 pay hike; the decline in the real value of MPs’ salaries that occurred in this period set in over the course of several years, not all at once just prior to 1962; the same is true of the increases in lawyers’ salaries that occurred in this period.[35] The only features of figure 5 that are consistent with the argument that higher pay for MPs attracts well-paid (and implicitly highly qualified) individuals to enter politics are the two slight increases in the ratio of lawyers to teachers among non-incumbent candidates that occur immediately after the introduction of salaries in 1953 and the pay hike of 1963. Equally, these increases were not only short-lived, they were small in magnitude relative to the sudden drop in the ratio of lawyers to teachers among candidates at the 1962 election, and they are statistically indistinguishable from random variation in the data.
5. Changes in educational attainment
Figure 6 shows the percentage of first-time MPs who possess an undergraduate university degree. We remind and caution the reader that this percentage is a minimum estimate, i.e., the percentage of MPs with undergraduate degrees is at least as high as shown in figure 6, but it may be higher. Recall also that we do not have a similar time series for parliamentary candidates, and as figure 5 above shows, there can be significant differences between MPs and candidates. MPs were always substantially better educated than the general adult population. In the 1950s this was largely due to the fact that only a small fraction of the population had graduated from university. However, it remains true even today when upwards of 20 per cent of Canadian adults possess university degrees. The House of Commons has obviously been successful in recruiting educated individuals. What is less obvious is what role if any MPs’ salaries and pensions have played in recruiting these educated individuals. Even to the naked eye, it is clear that the salary hike of 2001 had little impact in this regard. Aside from the dip in the percentage of degree-holders among first-time MPs in 1997, the percentage of first-time MPs with degrees was fairly stable from 1984 onward. Certainly, the percentage of new MPs with degrees did not rise after the 2001 pay hike (though it was already at a very high level). Much the same can be said with regard to the introduction of salaries and pensions in 1953; it had little immediate impact.
The 1963 salary increase stands out as quite different. The percentage of degree-holders among MPs elected at the 1965 election, the first general election to follow the pay increase, was 11 per cent higher than among those elected in 1963. Part of this increase was due to an increase in the number of university graduates in the adult population—but only a small part: the percentage of degree-holders among the adult population increased from 2.9 to 4.8 per cent between 1961 and 1971. One might contend that the sharp increase in degree-holding MPs at the 1965 election was merely the continuation of a trend that began at the 1962 election, when the percentage of first-time MPs with degrees jumped from 39.5 to 50.5 per cent. There is also a second surge in degree-holding first-time MPs in 1974, years after the 1963 salary increase to account for. Yet the continuation of the 1962 trend was not inevitable. In fact, it stalled almost immediately as the percentage of new MPs with degrees at the 1963 election fell slightly to 47.5. Similarly, the surge in 1974 tells us only that the 1963 salary increase was not a necessary condition for attracting degree-holders to the House of Commons; it certainly appears to have been sufficient condition, however.
Figure 6: Percentage of first-time MPs with undergraduate university degree or higher
Sources: 1949–1989: Derived from Historical Statistics of Canada, Degrees awarded by Canadian universities and colleges, by sex, Canada, selected years, 1831 to 1973 (Table W504-512), Ottawa: Statistics Canada, 2008 (www.statcan.gc.ca/pub/11-516-x/11-516-x1983001-eng.htm). 1990–2008 HRSDC calculations based on Statistics Canada. Labour force survey estimates (LFS), by educational attainment, sex and age group, annual (CANSIM Table 282-0004). Ottawa: Statistics Canada, 2012; Members of the House of Commons (www.parl.gc.ca/parlinfo/Lists/Members.aspx).
IV. Discussion
The introduction in 1953 of salaries and pensions for MPs, and the subsequent increases and alteration of these benefits in 1963 and 2001, were policies that were intended to attract a ‘better grade’ of candidate to the House: young, well-educated, and professional. We have attempted to assess the effectiveness of these policies. Our evidentiary standard has not been overly demanding: all we required to declare these policies as having achieved their objective was to observe a change in the age distribution, professional composition, or educational attainment of candidates or first-time MPs independent of what we might expect given contemporaneous changes in the broader population. Despite this lenient evidentiary standard, we find only one unambiguously positive result. The 1963 salary increase appears to have been sufficient to attract a cohort of well-educated first-time MPs to the House. Otherwise our results are negative. There is no evidence that higher salaries or pensions induced younger individuals to stand for office. Whilst the introduction of salaries and pensions in 1953 and the salary increase of 1963 coincided with an increase in the number of candidates drawn from law (a highly paid field) relative to the number of candidates drawn from teaching (a modestly paid field), the increases were small in magnitude, short-lived, and statistically indistinguishable from random variation in the data. There was also no indication that the changes in MPs’ remuneration effected in 1953 and 2001 had any impact on the educational attainment of first-time MPs.
Perhaps it could be argued that these negative results come about because the salary and pensions increases effected in 1953, 1963, and 2001 were not sufficient to raise MPs’ remuneration to a level high enough to attract and retain highly qualified individuals. We think this is a difficult counterargument to drive home. Figure 1 shows that MPs’ salaries have kept pace with inflation over the long run. Figure 3, moreover, shows that if MPs’ salaries no longer place them in the top one per cent of Canadian income earners, they still place MPs among the top five per cent (and this without accounting for the present value of their pensions). Figure 3 also makes clear that MPs are paid as well as all but the best-paid lawyers (see note 6) and substantially better than teachers. MPs’ remuneration has also kept level (albeit just) with the increased demands of the job as measured by the duration of parliamentary sessions. Indeed, the claim that the value of MPs’ remuneration has eroded substantially over time rests almost entirely on the premise that the non-accountable tax-free allowances that MPs received between 1945 and 2000 did not count as income. We reject this premise, not least because the Blais Commission itself rejected it.
What does this tell us about the desirability and effectiveness of using monetary compensation to motivate and control elected representatives? The argument that higher pay is necessary to attract highly talented or qualified individuals to enter politics does not appear to hold in the Canadian case. This might well be because the level at which Canadian MPs are remunerated is sufficiently high that additional increases generate only small marginal effects. This is probably the case with respect to educational attainment, for example, where more than two-thirds of new MPs hold undergraduate degrees and there is correspondingly less room for further increases. This case is harder to make with respect to the average age of candidates, however, because that has remained stubbornly high. Indeed, the evidence strikes us as quite consistent with the view that many candidates are drawn to politics, not for monetary but for ideological, personal, or altruistic reasons.[36] This does not imply that one could slash Canadian MPs’ salaries and pensions without effect. However, we should not worry that temporary declines in the economic value of MPs’ remuneration will translate directly into a decline in the quality of MPs. Further, we should desist in justifying MPs’ salaries and pensions on the need to attract talent to the House. A far stronger justification for paying MPs sufficiently (and that may be less or more than they currently receive) is that a certain level of income for MPs is necessary to prevent the House from becoming a plutocratic rather than a representative institution.
Question — Is your data looking at the upper house and the lower house or just the lower house? It is my understanding that in Canada you have got almost tenure in the upper house, so if you are looking for a job with job security you would naturally want to go to the upper house and not the lower house and that might skew the results. Why do you think the results are so different between Australia and Canada, given that remunerations are quite similar in their amounts?
Christopher Kam — The Senate in Canada is appointed, so whilst one might wish to go to the Senate, it is the prime minister who must ask or invite you to go. I have some data that I am working with, that shows you that senators stay about four terms. Moreover, even senators who are defeated MPs will stay that long. That tells you that there is a significant pent-up demand of a desire to stay in politics.
Why is the situation different in Australia? I think that begs a question about the different socio-economic and historical structure of our electoral politics. The fact of the matter is in Canada we have massive electoral swings and people are not deeply attached to parties. In Australia, both those conditions are not true and so you get much less volatility in Australian elections with the result that you have many more safe seats. We have maybe 25 or 30 per cent of seats in Canada that are safe and we have swings of 20 per cent.
Question — Have you thought to include the recruitment strategies of leaders of parties and the overall mix of professional and educational qualities as an explanatory factor? Second, what about implicit bargains for people who run or stay in office such as other career opportunities and appointments that only come through connections with the party? And thirdly, what about the strength of the party apparatus? For example, if you look at the Australian Labor Party it is very well organised which varies tremendously from Canada.
Christopher Kam — Those are great questions. Let me say that I have not talked to party leaders yet about their recruitment strategies but we do have some recent data from 2008 that tell us that 50 per cent of candidates were not invited by anybody. Here is the reality of Canadian politics: most constituency associations are not doling out a scarce good, they are desperate for warm bodies to put in front of the electorate, so they have to take what comes to them.
Post-parliamentary careers are much more of a phenomenon in Britain as an institution—where even backbenchers will get them—than in Canada. So we have some people of course such as Jean Chrétien or Marc Lalonde who go to the legal firm McCarthy Tétrault and have very good jobs. This is not true, as many commissions who have investigated the matters have found, of the vast majority of MPs who languish in two years of unemployment.
Your observation on the strength of parties to secure these kinds of goods or post-parliamentary careers is, I think, a very good one and it strikes me that one of the things that we have been very good at in Canada is killing parties. We absolutely killed the Conservative Party of Canada and it took fifteen years to reinvent itself. We pretty much drove a stake through the Liberal Party at the last couple of elections and so parties as organisations are incredibly weak, they are more franchise organisations rather than hierarchical corporate structures in Canada.
* This paper was presented as a lecture by Christopher Kam in the Senate Occasional Lecture Series at Parliament House, Canberra, on 22 March 2013. It was written while the first author was a visiting scholar in the School of Politics and International Relations at the Australian National University.
[1] Charles E.S. Franks, The Parliament of Canada, University of Toronto Press, Toronto, 1987, p. 75.
[2] Mathias Messner and Mattias K. Polborn, ‘Paying politicians’, Journal of Public Economics, vol. 88, no. 2–3, 2004, pp. 2423–45.
[3] Andrea Mattozi and Antonio Merlo, ‘Mediocracy’, National Bureau of Economic Research Working Papers, no: 12920, 2007, www.nber.org/papers/w12920. See also, Panu Poutvaara and Tuomas Takalo, ‘Candidate Quality’, International Tax and Public Finance, vol. 14, no. 1, 2007, pp. 7–27.
[4] To see how this can happen consider a dishonest but very competent politician. The politician’s competence provides her with lucrative opportunities in the outside economy, and hence salary increases are correspondingly less effective at constraining her behaviour. Voters may be unwilling to pay the very high salary necessary to keep this politician on her best behaviour, and if this is the case, voters will have to either tolerate the incumbent’s dishonesty or make do with a less competent (but perhaps more honest) alternative. Timothy Besley, ‘Paying politicians: Theory and evidence’, Journal of the European Economic Association, vol. 2, no. 2–3, 2004, pp. 193–215; Francesco Caselli and Massimo Morelli, ‘Bad Politicians’, Journal of Public Economics, vol. 88, no. 3–4, 2004, pp. 759–82.
[5] Claudio Ferraz and Frederico Finan, ‘Motivating politicians: The impacts of monetary incentives on quality and performance’, IZA Discussion Papers, no. 3411, http://hdl.handle.net/10419/35111.
[6] Michael M. Atkinson and Dustin Rogers, ‘Better politicians: If we pay, will they come?’, presented at the Annual Meeting of the Canadian Political Science Association, Edmonton, 2012, www.cpsa-acsp.ca/papers-2012/Atkinson.pdf.
[7] Kaisa Ilona Kotakorpi and Panu Poutvaara, ‘Pay for politicians and candidate selection: An empirical analysis’, CESifo Working Paper Public Choice, no. 3126, http://hdl.handle.net/10419/39019.
[8] Raymond Fisman, Nikolaj A. Harmon, Emir Kamenica and Inger Munk, ‘Labor supply of politicians’, National Bureau of Economic Research Working Paper, no. 17726, www.nber.org/papers/w17726.
[9] ibid.
[10] Atkinson and Rogers, op. cit.
[11] Franks, op. cit.
[12] Michael M. Atkinson and David C. Docherty, ‘Moving right along: The roots of amateurism in the Canadian House of Commons’, Canadian Journal of Political Science, vol. 25, no. 2, 1992, pp. 295–318.
[13] V. S. Harder, ‘Career Politics and Political Parties’ in J-P Gaboury and J. R. Hurley (eds), The Canadian House of Commons Observed: Parliamentary Internship Papers, University of Ottawa Press, Ottawa, 1979.
[14] Franks, op. cit.
[15] Anthony King, ‘The rise of the career politician in Britain – and its consequences’, British Journal of Political Science, vol. 11, no. 3, 1981, pp. 249–85.
[16] Max Weber, ‘Politics as a vocation’ in H.H. Gerth and C. Wright Mills (eds.), From Max Weber: Essays in Sociology, Routledge and Kegan Paul, London, 1948, p. 84.
[17] King, op. cit., p. 250.
[18] Prime Minister Louis Stephen St. Laurent, House of Commons Debates (Hansard), 25 June 1952, p. 3678.
[19] Indeed, in a 1981 debate on MPs’ pay, Walter Baker wryly observed that any argument that higher salaries were needed to attract better people to the House was an admission that the current membership left a lot to be desired. Government House Leader Walter Baker, House of Commons Debates (Hansard), 9 July 1981, p. 11395–6.
[20] Member of Parliament Agar Rodney Adamson, House of Commons Debates (Hansard), 25 June 1952, p. 3687.
[21] Government House Leader Rt. Hon. Mitchell William Sharp, House of Commons Debates (Hansard), 17 December 1974, p. 2333.
[22] Official Opposition House Leader Gerald William Baldwin, House of Commons Debates (Hansard), 17 December 1974, p. 2337.
[23] Canada. Advisory Committee on Parliamentary Salaries and Expenses (Beaupré Committee), Report of the Advisory Committee on Parliamentary Salaries and Expenses, [Ottawa], 1970, p. 39.
[24] Canada. Commission to Review Allowances of Members of Parliament (Blais Commission), Supporting Democracy, [Ottawa], 1998, vol. 1, p. 15. The commission was chaired by Jean-Jacques Blais.
[25] Max Saltsman (in the 1952 debate) and Stanley Knowles (in a 1981 debate on MPs’ pay) were significant exceptions to this generalisation. Knowles disputed the argument that higher salaries would attract more talented or better qualified individuals to seek election to the House. The only thing that could be guaranteed, Knowles argued, was that higher salaries would attract people interested in higher salaries (House of Commons Debates (Hansard), 9 July 1981, pp. 11395–6).
[26] We estimate the taxable equivalent by multiplying the base amount of the tax-free allowance ($22 500) by a factor of 1.4. This scaling factor is based on the results of a comparative study of Canadian legislative salaries conducted by the British Columbia legislature (see www.leg.bc.ca/bcmlacomp/media/PDF/Remuneration-Comparisons.pdf). Clearly, this is just an approximation and in reality the scaling factor would vary by an MP’s province of residence and what tax bracket they fell into given any outside earnings, tax shelters etc.
[27] Blais Commission, Supporting Democracy, op. cit., vol. 1, p. 32.
[28] All sessions after 1953 (when salaries and pensions were introduced) were assumed to last the entire year (i.e. they were weighted by a factor of one).
[29] Blais Commission, Supporting Democracy, op. cit., vol. 2, pp. 91–2.
[30] ibid., vol. 1, p. 32. Two other factors also contribute to the discrepancy between our results and those of the Blais Commission:
1. The Blais report based its comparison of MPs’ and teachers’ salaries on the salaries of secondary school teachers. In contrast, our comparison is based on the salaries of primary and secondary school teachers. Our decision on this matter was driven by the fact that the census did not distinguish between primary and secondary school teachers until 1971. The implications of this decision are not trivial. Secondary school teachers earn significantly higher salaries than primary school teachers. Moreover, even as late as 2005 female teachers earned substantially less than male teachers. Not only do primary school teachers outnumber secondary school teachers 2:1, but primary school teachers are disproportionately female. The upshot is the Blais Commission was essentially comparing MPs’ salaries to those of the very highest earning teachers.
2. The census and labour force surveys from which our data are derived provide estimates of lawyers’ salaries. This raises two issues. First, lawyers who are employed by larger firms may earn significant annual bonuses that do not appear in our data. Second, lawyers who are paid mainly via salary tend to be either associates or in-house councils. In contrast, senior partners may be remunerated by other means (e.g., by taking a dividend of the firm’s profits). We are therefore sensitive to the possibility that our own comparison of MPs’ and lawyers’ earnings is based on an underestimate of lawyers’ earnings. Still, looking at industry reports (e.g., Robert Half Legal 2011, 18 [www.law.ca/system/files/RHL_SalaryGuide_2011%5B15D.pdf]) shows that MPs’ earnings are commensurate with those of lawyers of 4–9 years’ experience at mid- and large-sized Canadian law firms. Only the most senior lawyers at the largest firms earn substantially more than MPs.
[31] We limit our attention to first-time MPs for the logical reason that the characteristics of sitting MPs cannot be due to policies that were passed after they were elected.
[32] The database can be accessed at www.parl.gc.ca/parlinfo/.
[33] The average age of non-incumbent candidates declined by 2 years and 2 months, from 49.8 years to 47.6 years. A t-test indicates that this difference is statistically significant at p <.05.
[34] It could, of course, be the case that lawyers’ propensity to enter or leave politics is driven more by changes in lawyers’ salaries rather than by changes in the remuneration of MPs. Such a dynamic could result in lawyers retreating from politics even as the remuneration of MPs increases. Even if this were the case, however, the ratio of lawyers to teachers among parliamentary candidates would still provide an accurate estimate of the relative willingness of lawyers and teachers to enter politics.
[35] Moore describes a terrific growth in the demand for and remuneration of legal services starting in the late-1940s and lasting through to the early 1970s. Much of the initial growth, Moore argues, was driven by the adoption of the Rand formula (governing industrial relations) in 1944 and significant overhauls of the tax code in 1949 and 1952. The reader will note that all this legislation pre-dates the sharp and sudden decline in 1962 of candidates drawn from the legal profession. See Christopher Moore, The Law Society of Upper Canada and Ontario’s Lawyers 1797–1997, University of Toronto Press, Toronto, 1997, chapter 5.
[36] See, for example, Pippa Norris and Joni Lovenduski, Political Recruitment: Gender, Race and Class in the British Parliament, Cambridge University Press, Cambridge, 1995.
Prev | Contents | Next